People who are actively engaging in nonsuicidal self-injury, defined as five or more episodes within the past year, score significantly higher on the personality traits of negative affectivity and psychoticism than psychiatric patients whose self-injury lies in the past or who have never self-injured, according to a cross-sectional study of 231 adults treated at general psychiatric outpatient units in Lund, Sweden. The finding, published in the journal Borderline Personality Disorder and Emotion Dysregulation, suggests that the traits distinguishing ongoing self-harm from recovered or absent self-harm are narrower than many researchers had assumed.
The study, registered prospectively on ClinicalTrials.gov in April 2021 and approved by the Swedish Ethical Review Authority, set out to answer a question that earlier work had largely sidestepped. Most previous investigations compared people who self-injure with healthy controls, which makes it impossible to tell whether an elevated trait is specific to self-injury or simply a feature of mental illness in general. By recruiting all participants from clinical populations and dividing them by the timing of their self-injury, the Swedish team could ask which maladaptive personality traits track the persistence of the behavior rather than psychopathology broadly.
Participants were enrolled between April 30, 2021, and August 26, 2022, from two general psychiatric outpatient care units. All were at least 18 years old, and the final sample ranged in age from 18 to 35. Patients with a registered main diagnosis of psychotic disorder or intellectual disability were excluded from recruitment, and anyone who had ever received such a diagnosis, primary or secondary, was removed during screening. After further exclusions for language difficulties, lack of ongoing treatment, and missing data, 231 participants entered the final analyses. Each received the equivalent of about 35 euros in compensation.
The researchers grouped participants using the Inventory of Statements About Self-injury, a questionnaire covering lifetime frequency of twelve different forms of self-harm, supplemented with questions about how many episodes had occurred in the last three months and the last year. The active group comprised 123 individuals with five or more episodes in the past twelve months, matching the DSM-5 Section III research criteria for nonsuicidal self-injury disorder. The previous group comprised 68 individuals with a comparable lifetime history but no episodes in the past year. The remaining 40 participants had no lifetime history of self-injury or at most four episodes.
Personality was measured with the Personality Inventory for DSM-5, a 220-item self-report questionnaire developed for the Alternative Model for Personality Disorders introduced in DSM-5. It yields scores on five maladaptive trait domains: negative affectivity, detachment, antagonism, disinhibition, and psychoticism. Internal consistency in this sample was good to excellent, with Cronbach’s alpha ranging from 0.89 to 0.94 across domains. Because self-reported personality can be colored by current mood, the team also administered the GAD-7 anxiety scale and the self-report version of the Montgomery-Åsberg Depression Rating Scale, and entered both, along with age and sex assigned at birth, as covariates in the analyses.
The central result concerned two of the five domains. Negative affectivity, which in the DSM-5 framework captures emotional instability and proneness to intense negative emotions, differed significantly across the three groups, with a medium effect size. Adjusted comparisons showed that the active group scored significantly higher than both the no-history group and the previous-self-injury group, while the previous and no-history groups did not differ from each other. Psychoticism showed a parallel pattern: significantly higher in the active group than in those with no history, and close to significantly higher than in those with past self-injury, again with no difference between the latter two groups. Importantly, these differences survived adjustment for depressive and anxiety symptoms, indicating that they were not merely an artifact of current symptom burden.
The remaining three domains told a different story. Antagonism, detachment, and disinhibition showed no significant associations with self-injury status in the primary analyses. This contrasted with earlier studies, largely conducted in non-clinical student samples, which had reported elevated antagonism and disinhibition among people who self-injure. The authors suggest that those earlier elevations may reflect broader psychopathology rather than characteristics specific to self-injury. Sensitivity analyses that dropped the depression and anxiety covariates did shift the picture somewhat: detachment then emerged as significantly higher in the active group, and disinhibition showed a difference at the conventional threshold that did not survive the study’s stricter Bonferroni-corrected criterion, underscoring how sensitive some domain findings are to the handling of current symptoms.
The elevated psychoticism scores deserve particular emphasis, because within the Alternative Model for Personality Disorders the domain does not imply a psychotic illness. It also captures perceptual dysregulation, including depersonalization, derealization, and dissociative experiences. Dissociation has been closely linked to self-injury, and the authors note that these traits may reflect difficulties mentalizing under emotional stress, verbalizing internal experiences, and maintaining a coherent sense of self and others, impairments that could feed into affect dysregulation and increase vulnerability to self-harm as a maladaptive emotion-regulation strategy. Only one prior study, in young adults attending youth mental health services, had reported elevations in both negative affectivity and psychoticism, and that study also found detachment and antagonism differences that did not replicate here.
The cohort itself carried a heavy clinical burden. Depression and anxiety scores were highest in the active group, and suicide attempts, extracted from medical records rather than self-report, ranged from zero to 45 across the sample, with the active group showing the highest mean and the widest range, including two extreme outliers of 42 and 45 attempts. Twenty-two participants had a history of suicide attempts requiring somatic medical care, fourteen of them in the active group. The active and previous groups were predominantly female, whereas the no-history group was evenly divided, and the distribution of current psychiatric diagnoses, dominated by anxiety and stress-related disorders, attention-deficit/hyperactivity disorder, and mood disorders, did not differ significantly across groups.
The authors are careful about what the design can and cannot show. As a cross-sectional study, it cannot establish whether the elevated traits precede and contribute to persistent self-injury or arise alongside it, and it relied on self-reported timing and frequency of self-injury, a known source of recall bias; one cited comparison of self-report against medical records in an Australian prison found less than 40 percent accuracy. The sample also excluded people over 35 and showed a sex imbalance, and the measures assessed constructs over mismatched timeframes, with personality treated as stable while symptoms and behavior were anchored to recent weeks. The absence of a measure of personality functioning means general severity of psychopathology cannot be ruled out as an explanation. What the study does establish is that, within a psychiatric population and after adjustment for age, sex, depression, and anxiety, ongoing self-injury is specifically marked by heightened negative emotionality and psychoticism-like perceptual dysregulation, a pattern the authors hope will guide future longitudinal work on the mechanisms, course, and treatment of persistent self-harm.
Subject of Research: Maladaptive personality traits associated with active versus previous nonsuicidal self-injury in psychiatric outpatients
Article Title: Elevated Negative Affectivity and Psychoticism in individuals with active Nonsuicidal Self-Injury (NSSI) compared with individuals with previous or no NSSI
Article References: Thylander, C., Nilsson, M., van Westen, D., Björkstrand, J., & Westling, S. (2026). Elevated Negative Affectivity and Psychoticism in individuals with active Nonsuicidal Self-Injury (NSSI) compared with individuals with previous or no NSSI. Borderline Personality Disorder and Emotion Dysregulation, 13(1), Article 28. https://doi.org/10.1186/s40479-026-00373-z
Image Credits: AI Generated
DOI: 10.1186/s40479-026-00373-z
Keywords: nonsuicidal self-injury, NSSI, personality traits, PID-5, negative affectivity, psychoticism, DSM-5, Alternative Model for Personality Disorders, psychiatric outpatients, cross-sectional study, dissociation, suicide attempts
News Source: Glenn Wilkins. (October 29, 2026). Active Self-Harm Linked to Higher Negative Affectivity and Psychoticism in Psychiatric Patients. Scienmag.
